Method for on-line film thickness detection of a cavity wax injection system

Abstract

This invention relates to the field of thickness detection technology, specifically disclosing an online film thickness detection method for a cavity wax injection system. The method requires a detection system comprising an image acquisition unit, an image processing unit, a driver, and a control unit. The detection steps are as follows: the control unit checks the interlock signals between the driver, image acquisition unit, guide joint, and image processing unit to ensure they are OK, and obtains relevant position information; then, the control unit controls the image acquisition unit to acquire images; subsequently, the image processing unit compares the real-time image with standard images in a film thickness database to obtain measured film thickness data; finally, the control unit compares the measured film thickness data with preset allowable data and outputs an NG signal and issues an alarm signal when the measured film thickness data exceeds the preset allowable data. This solution solves the problems of low efficiency, high cost, and inability to perform online detection during product manufacturing associated with existing manual inspection methods for cavity wax injection.

Description

[0001] This application is a divisional application of the patent application filed on November 25, 2021, with application number 2021115539109, entitled "Online Film Thickness Detection System for Cavity Wax Injection System". Technical Field

[0002] This invention relates to the field of thickness detection technology, specifically to an online film thickness detection method for a cavity wax injection system. Background Technology

[0003] In recent years, national standards have required that the body-in-white have a lifespan of 10 to 15 years without perforation corrosion. There are many cavities in the car body designed to increase body strength and reduce body weight. These cavities are very complex. During the rust prevention treatment of the body-in-white using "cathodic electrophoresis coating", the electrophoretic film inside the cavity is very thin due to the Faraday effect, and the cavity surface is usually prone to rust. In actual production, the method of wax injection into the body cavity is usually used to achieve corrosion prevention.

[0004] Existing cavity wax injection methods are mainly divided into four types: traditional wax injection, semi-automatic wax injection, fully automatic wax injection, and pouring wax. Among them, the pouring wax process is rarely used due to its high energy consumption and equipment investment. The other three methods are all carried out by spraying. In the traditional wax injection, semi-automatic wax injection, and fully automatic wax injection processes, the effective coverage of the wax film on the cavity is the key to achieving corrosion prevention.

[0005] The existing wax injection process only includes the spraying process. The actual coverage effect of the wax film is mainly confirmed by the inspection of the prototype vehicle or sample to determine the spraying effect of the wax film thickness. The inspection of the prototype vehicle or sample is carried out manually to confirm the spraying effect. There are usually two methods for manual inspection. One is to process additional observation holes on the prototype vehicle or sample so that the manual inspection can directly check the spraying condition inside the cavity. The other is to use the existing wax injection hole of the cavity and use an endoscope to check the cavity spraying condition in real time.

[0006] To facilitate manual inspection, a fluorescent agent is added to the wax to be sprayed before the cavity is coated. After coating, when the cavity surface is illuminated with a UV lamp, the areas of the wax film containing the fluorescent agent will fluoresce. Manual inspection of the fluorescence on the inner surface of the cavity allows for judgment of whether there is insufficient wax or an excessively thin wax film, thus confirming the sample's quality. While this method can inspect the wax filling process in the cavity, it still has the following problems:

[0007] First, when manually inspecting the wax spraying of cavities directly through the observation hole, there is usually only one observation hole. Observing the spraying of all surfaces inside the cavity through a single observation hole presents blind spots due to the complex cavity structure, making it impossible to complete a comprehensive inspection of the cavity surfaces and reducing the accuracy of the cavity wax injection effect inspection. On the other hand, whether it is a direct inspection through the observation hole or an inspection using an endoscope, the complex cavity structure and the large number of cavities in the car body result in long inspection times and high labor costs for each inspection. Therefore, in actual production, most inspections are carried out by sampling. However, the randomness of sampling further reduces the accuracy of the inspection.

[0008] Secondly, when manually inspecting the wax injection in the cavity, since the spraying is done by the naked eye, the inspection process can only see whether there is insufficient wax, excessively thick or thin wax film in some areas. If there is insufficient wax, it means that there is a missed spraying during the wax injection process, and it can be directly judged as a defective product. However, if there is a case of excessively thin wax film, it is difficult to determine whether the wax film thickness is up to standard by the naked eye, which increases the difficulty of inspection. Summary of the Invention

[0009] The present invention aims to provide an online film thickness detection system for cavity wax injection systems, in order to solve the problems of reduced accuracy, low efficiency and high labor costs associated with the existing method of manually inspecting the cavity wax injection through observation holes.

[0010] To achieve the above objectives, the basic solution of the present invention is as follows:

[0011] An online film thickness detection system for cavity wax injection systems includes an image acquisition unit and an image processing unit. The image processing unit is connected to the image acquisition unit and also includes a driver. The image acquisition unit is connected to the output of the driver. The driver can drive the image acquisition unit to extend into the wax injection cavity. The image acquisition unit can acquire images of different areas inside the cavity. The image processing unit has a film thickness database. The image processing unit receives the real-time images acquired by the image processing unit and compares the received real-time images with standard images in the film thickness database to obtain the measured film thickness data corresponding to the real-time images.

[0012] Compared to existing technologies, the following benefits are achieved:

[0013] When using this solution, during the inspection of the waxing effect of a workpiece after cavity waxing, the driver automatically drives the image acquisition unit into the waxing cavity. After entering the cavity, the image acquisition unit acquires images of different areas on the cavity surface. The image processing unit compares multiple images acquired in real time within the cavity to obtain measured film thickness data for different areas within the cavity. The measured film thickness data allows for a quick determination of whether the coating effect of the waxing cavity is qualified. Compared with existing technologies, this solution makes the measured film thickness data within the cavity more intuitive, avoiding the difficulty of manual inspection. Furthermore, the intuitiveness of the measured film thickness data facilitates subsequent analysis of the film thickness data.

[0014] Furthermore, when using this solution, the image acquisition unit can quickly extend into the cavity to acquire multiple images under the drive of the driver. The operation speed is fast, and the ability of the image acquisition unit to extend into the cavity avoids the problem of reduced detection accuracy caused by blind spots. At the same time, this solution can acquire multiple images in a short time through the image acquisition unit, which makes it convenient for the image processing unit to quickly determine the actual thickness of the membrane inside the cavity based on the acquired images, so as to determine whether the cavity wax injection is qualified. This improves the inspection efficiency of cavity wax injection, facilitates the replacement of manual inspection by this system, and enables online detection of wax-injected cavities. It achieves the goal of improving detection efficiency and inspection accuracy while reducing labor costs.

[0015] In addition, existing methods for detecting coating thickness, such as ultrasonic, magnetic induction, eddy current, magnetic pulse, and photothermal methods, have limitations. These limitations include the complex structure of the wax injection cavity, the presence of small internal curved surfaces within the cavity, and the small and unevenly distributed size of the wax injection holes connecting the cavity to the outside. Ultrasonic, magnetic induction, and eddy current methods cannot achieve online detection of the wax injection cavity, while magnetic pulse methods can only detect the film thickness at a limited number of points online, resulting in incomplete film thickness detection. Photothermal methods are not suitable for complex internal surfaces. In contrast, this system can meet the requirements for detecting the coating effect of the wax injection cavity, obtain measured film thickness data for different areas of the wax injection cavity, and achieve high detection speed for online detection.

[0016] Furthermore, it also includes a control unit, which is connected to the image processing unit. The control unit is used to receive the measured film thickness data obtained by the image processing unit, and to compare the received measured film thickness data with preset allowable data and issue an alarm signal when the measured film thickness data exceeds the preset allowable data.

[0017] Beneficial effects: This solution automatically determines the measured film thickness data through the setting of the control unit, which improves the automation level of this detection system. At the same time, since this detection system can perform real-time online detection after cavity wax injection, the alarm information issued by the control unit of this detection system can remind workers to quickly adjust the wax injection situation, realize online monitoring, reduce the probability of batch wax injection defects, and enable rapid feedback on the wax injection effect, which is conducive to improving the wax injection qualification rate and reducing production costs.

[0018] Furthermore, the control unit is connected to the image acquisition unit, and the control unit is used to control the image acquisition unit to perform image acquisition, so as to improve the automation level of image acquisition.

[0019] Furthermore, a guide joint is connected between the output end of the driver and the image acquisition unit. The guide joint is used to drive the image acquisition unit to form different angles relative to the output end of the driver.

[0020] Beneficial effects: By setting the guide joint, the angle of the image acquisition unit relative to the output end of the driver can be adjusted, thereby changing the shooting angle of the image acquisition unit to ensure that the acquired cavity image is more comprehensive and further improve the accuracy of cavity wax injection detection.

[0021] Furthermore, the control unit is connected to the guide joint, and the control unit controls the angle formed between the image acquisition unit and the output end of the driver by controlling the swing angle of the guide joint.

[0022] Beneficial effects: By controlling the guide joint through the control unit to drive the image acquisition unit to acquire images at different angles, the degree of automation is further improved.

[0023] Furthermore, the control unit is connected to the driver, and the control unit is used to control the driver to drive the image acquisition unit to extend into or withdraw from the wax injection cavity, so as to accurately deliver the image acquisition unit connected to the output end of the driver into the cavity through automatic control.

[0024] Furthermore, the image acquisition unit includes an optical lens, an illumination source, and an ultraviolet light source.

[0025] Beneficial effects: The illumination source and ultraviolet light provide the conditions for image acquisition, resulting in higher image quality captured by the optical lens.

[0026] Furthermore, the film thickness database includes several standard images corresponding to changes in film thickness while keeping the proportion of fluorescent agent added to the spray wax and the shooting angle constant, as well as several standard images corresponding to changes in the proportion of fluorescent agent added to the spray wax while keeping the shooting angle and film thickness constant.

[0027] Beneficial Effects: When performing wax cavity film thickness testing on a car body using this solution, the shooting angle is typically set so that the lens is perpendicular to the cavity surface being captured, ensuring a fixed shooting angle. Combining this solution's single-variable principle, the proportion of fluorescent agent added to the wax and the film thickness are used as variables to obtain several standard images, forming a film thickness database. This database of standard images ensures that the obtained standard images are independent of the cavity structure, relying instead on the proportion of fluorescent agent added to the wax and the film thickness. This allows for easier processing even after vehicle model changes; only the proportion of fluorescent agent added to the wax needs to be known. The image processing unit can then find a matching standard image based on the real-time image and obtain the measured film thickness data corresponding to the real-time image. Therefore, the film thickness database established by this solution does not require capturing different images based on the specific geometry of each cavity using the single-variable principle, significantly reducing the workload of database creation. It also avoids the problem of inconsistent acquisition conditions when capturing images of different cavities at different times, which reduces the reliability of the acquired standard images, thus improving the accuracy of the image processing unit.

[0028] Furthermore, the film thickness database also includes several standard images corresponding to the proportion of fluorescent agent added to the sprayed wax and the change in shooting angle while the film thickness remains unchanged. The image processing unit can import the geometric data of the cavity and can determine the shooting angle when the image acquisition unit acquires the image based on the geometric data of the cavity and the position of the image acquisition unit in the cavity.

[0029] Beneficial effects: To further enhance the intelligence of this detection system, cavity geometry data is introduced into the image processing unit. The film thickness database is expanded to include the proportion of fluorescent agent added to the sprayed wax and several standard images corresponding to varying shooting angles while maintaining a constant film thickness. This allows the image processing unit to automatically identify and confirm the shooting angle simply by knowing the position of the image acquisition unit extended by the driver, the position and angle of the lens on the image acquisition unit, and the geometric model of the cavity being inspected, thus improving the intelligence level of the detection system. Furthermore, since the standard images in the film thickness database are taken under three different variables, the same image may correspond to different variables. This solution automatically identifies the shooting angle, facilitating the determination of the standard images to be compared with the real-time image, thereby improving the accuracy of the image processing unit's judgment.

[0030] Furthermore, before comparing the real-time image with the standard image in the film thickness database, the image processing unit needs to perform image preprocessing on the received real-time image. Image preprocessing includes denoising and enhancement of the real-time image. The image processing unit uses image processing algorithms to segment regions with different textures and colors from the background into non-overlapping regions. The image processing unit imposes conditional constraints on the segmented regions, including at least one of area constraints, perimeter constraints, and shape constraints. Regions that meet all conditional constraints are valid feature regions. Finally, the image processing unit compares the obtained valid region features with the standard image in the film thickness database to obtain the measured film thickness data corresponding to the real-time image.

[0031] Beneficial Effects: In practical applications, due to the complex internal structure of cavities, it is difficult to ensure that the cavity surfaces of the acquired images are on the same plane. This results in differences in texture or color in different regions of the acquired real-time images. If only a fixed region in the image is compared with a standard image, the comparison results may be inaccurate. To solve this problem, this solution uses an image processing unit to preprocess the received real-time images by denoising and enhancing them. This eliminates irrelevant information such as noise, impurities, spraying defects, or dust, and restores and enhances real information, such as enhancing the contrast between the sprayed area color and the background color, thereby improving the detectability of the obtained real-time images and simplifying the data to the greatest extent. Then, the preprocessed image is segmented according to an image processing algorithm to obtain non-overlapping regions. The segmented non-overlapping regions are then subject to constraints based on at least one of the following: area, perimeter, and shape. Regions that meet all constraints are considered valid feature regions in the real-time image. Finally, the multiple valid regions are compared with standard images in the film thickness database to obtain the measured film thickness data of different regions under the real-time image, thus improving the accuracy of the comparison results. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the main structure of Embodiment 1 of the present invention;

[0033] Figure 2 This is a connection diagram of Embodiment 1 of the present invention;

[0034] Figure 3 This is a flowchart of a detection method for detecting the film thickness in a wax-filled cavity using Embodiment 1 of the present invention. Detailed Implementation

[0035] The following detailed description illustrates the specific implementation method:

[0036] The reference numerals in the accompanying drawings include: image acquisition unit 1, guide joint 2, moving arm 3, and driver 4.

[0037] Example 1

[0038] Example 1 is basically as shown in the appendix. Figures 1 to 3 As shown, the online film thickness detection system for the cavity wax injection system includes a driver 4, an image acquisition unit 1, an image processing unit, and a control unit. The output end of the driver 4 is fixedly connected to a movable arm 3. A guide joint 2 is connected between the movable arm 3 and the image acquisition unit 1. The guide joint 2 is used to drive the image acquisition unit 1 to form different angles relative to the movable arm 3. The image acquisition unit 1 can extend into the wax injection cavity and can acquire images of different areas inside the cavity.

[0039] The image processing unit is connected to both the image acquisition unit 1 and the control unit. The control unit is also connected to the driver 4, the image acquisition unit 1, and the guide joint 2. The control unit is used to control the working time of the driver 4. In this embodiment, the driver 4 is a servo cylinder, and the moving arm 3 is fixedly connected to the piston rod of the servo cylinder. Of course, the driver 4 can also be other mechanisms or devices that can drive the image acquisition unit 1 to extend into the wax injection cavity, such as a robot, a linear motor, a three-dimensional moving slide, etc.

[0040] Image acquisition unit 1 includes an optical lens, an illumination source, and an ultraviolet light source, wherein the illumination source adopts an LED light source. Image acquisition unit 1, guide joint 2, and moving arm 3 are equivalent to forming a controllable endoscope. The specific structure of guide joint 2 may include the articulated movable section and the physical control rod for controlling the angle swing of the articulated movable section disclosed in the Chinese patent announcement number CN105408069B, patent name "Pipe Endoscope Manipulation and Adjustment System and Method". Through the connection of the physical control rod and the control unit, the angle swing of guide joint 2 is controlled, thereby controlling the angle change of image acquisition unit 1 connected to guide joint 2 relative to moving arm 3, so that image acquisition unit 1 can acquire different images at different angles. The control unit controls image acquisition unit 1 to perform image acquisition after the angle adjustment is completed.

[0041] The image processing unit is equipped with a film thickness database. The image processing unit receives real-time images acquired by the image processing unit and compares the received real-time images with standard images in the film thickness database to obtain the film thickness measurement data corresponding to the real-time image. The image processing unit stores the obtained film thickness measurement data and sends the film thickness measurement data to the control unit.

[0042] The control unit compares the received film thickness measurement data with the preset allowable data and issues an alarm signal when the film thickness measurement data exceeds the preset allowable data. In this embodiment, the control unit can be a PLC controller.

[0043] When measuring the film thickness in the cavity after wax injection, the detection method using this detection system is as follows:

[0044] Step A: The control unit checks whether the interlocking signal between the driver 4, image acquisition unit 1, guide joint 2, and image processing unit is OK, and obtains the relevant position information, including the wax injection hole position information and the image acquisition unit 1 position information, etc.

[0045] Step B: The control unit controls the driver 4 to drive the image acquisition unit 1 into the wax injection cavity until the image acquisition unit 1 moves to the set position; then the illumination source and ultraviolet light source of the image acquisition unit 1 are turned on to provide conditions for the optical lens to acquire images.

[0046] Step C: The control unit controls the guide joint 2 to swing to adjust the shooting angle of the image acquisition unit 1. The control unit controls the image acquisition unit 1 to perform image acquisition after each shooting angle adjustment.

[0047] Step D: The image acquisition unit 1 transmits the acquired image to the image processing unit in real time. The image processing unit compares the received real-time image with the standard image in the film thickness database to obtain the film thickness measurement data corresponding to the real-time image. The image processing unit stores the film thickness measurement data and sends the film thickness measurement data to the control unit.

[0048] Step E: The control unit compares the received film thickness measurement data with the preset allowable data, and outputs an NG signal and issues an alarm signal when the film thickness measurement data exceeds the preset allowable data.

[0049] The detection system in this embodiment can automatically detect the cavity after wax injection online, realizing online monitoring of the wax injection effect, enabling rapid feedback on the wax injection effect, which helps to improve the wax injection qualification rate and reduce production costs.

[0050] Furthermore, during the inspection process, the guide joint 2 and the driver 4 enable image acquisition of the inner surface of the cavity with complex structure, avoiding the problem of reduced inspection accuracy due to visual blind spots caused by manual inspection. At the same time, through image data acquisition and processing by the image processing unit, the measured film thickness data of different areas inside the cavity can be accurately obtained. This not only makes the measured film thickness data inside the cavity more intuitive, but also allows the control unit to quickly determine whether the coating effect of the wax-filled cavity is qualified, avoiding the problem of high inspection difficulty in manual inspection.

[0051] In addition, this detection system automates position movement before image acquisition, angle adjustment during image acquisition, and image capture. It also automates post-image acquisition processing and result judgment. It can replace manual labor to automatically detect the film thickness of wax injection cavities, greatly improving the level of automation and intelligence, saving labor costs, and increasing detection efficiency.

[0052] Example 2

[0053] Example 2 is based on Example 1 with the following changes: In this example, the film thickness database includes standard images collected according to the principle of a single variable, with the proportion of fluorescent agent added to the spray wax and the film thickness as variables. That is, the film thickness database includes several standard images corresponding to the film thickness changing while the proportion of fluorescent agent added to the spray wax remains unchanged, and several standard images corresponding to the film thickness changing while the shooting angle and film thickness remain unchanged.

[0054] When performing wax cavity film thickness detection on a car body using this embodiment, the shooting angle is generally set so that the shooting lens is perpendicular to the cavity surface of the image being captured, making the shooting angle a fixed value. Combined with the single-variable principle of this embodiment, the proportion of fluorescent agent added to the spray wax and the film thickness are used as variables to obtain several standard images, forming a film thickness database. By establishing this film thickness database containing several standard images, the obtained standard images do not depend on the cavity structure, but rather on the proportion of fluorescent agent added to the spray wax and the film thickness. This makes it convenient that after a change in car model, only the proportion of fluorescent agent added to the spray wax for that model needs to be known. The image processing unit can find a matching standard image based on the real-time image, and then obtain the measured film thickness data corresponding to the real-time image based on the film thickness data corresponding to the standard image. Therefore, the film thickness database established by this embodiment does not require taking different images based on the specific geometry of each cavity according to the single-variable principle, thus greatly reducing the workload of establishing the film thickness database. It also avoids the problem of inconsistent acquisition conditions when acquiring images of different cavities at different times, which reduces the reliability of the acquired standard images, and helps improve the accuracy of the image processing unit.

[0055] Example 3

[0056] Example 3 is a further improvement on Example 2, as follows: The film thickness database also includes several standard images corresponding to the ratio of fluorescent agent added to the sprayed wax and the change of shooting angle while the film thickness remains unchanged. The image processing unit can import the geometric data of the cavity and can determine the shooting angle when the image acquisition unit 1 acquires the image based on the geometric data of the cavity and the position of the image acquisition unit 1 in the cavity.

[0057] To further enhance the intelligence of this detection system, this embodiment introduces cavity geometry data into the image processing unit and adds several standard images corresponding to the varying shooting angles while maintaining a constant film thickness to the film thickness database, based on the proportion of fluorescent agent added to the sprayed wax. This allows the image processing unit to automatically identify and confirm the shooting angle by knowing only the position of the image acquisition unit 1 extended by the driver 4, the position and angle of the shooting lens on the image acquisition unit 1, and the geometric model of the cavity being detected, thus improving the intelligence level of this detection system. In addition, since the standard images in the film thickness database are taken under three different variables, the same image may correspond to different variables. This embodiment automatically identifies the shooting angle, thereby facilitating the determination of the standard images to be compared with the real-time image and improving the accuracy of the image processing unit's judgment.

[0058] Example 4

[0059] Example 4 is a further improvement on Example 1. Of course, Example 4 is not limited to improvements on Example 1; it could also be an improvement on Example 2 or Example 3. This Example 4 is based on Example 1, with the following specific improvements: Before comparing the real-time image with the standard image in the film thickness database, the image processing unit needs to perform image preprocessing on the received real-time image. Image preprocessing includes denoising and enhancement of the real-time image. The image processing unit uses an image processing algorithm to segment regions with different textures and colors from the background into non-overlapping regions. The image processing unit imposes conditional constraints on the segmented regions, including area, perimeter, and shape constraints. Regions that meet all conditional constraints are considered valid feature regions. Finally, the image processing unit compares the obtained valid region features with the standard image in the film thickness database to obtain the measured film thickness data corresponding to the real-time image. Furthermore, the image processing unit calculates the pixel area of ​​the valid feature region and, combined with the nonlinear calibration results of the optical lens, converts the pixel area of ​​the valid feature region into a physical area in millimeters, thereby determining and outputting the size of the valid feature region.

[0060] The specific process of this embodiment is as follows: This embodiment uses an image processing unit to perform denoising and enhancement operations on the received real-time image, thereby eliminating irrelevant information in the image such as noise, impurities, spraying defects or dust, and restoring and enhancing real information such as enhancing the contrast between the spraying area color and the background color, so as to improve the detectability of the obtained real-time image and simplify the data to the maximum extent. Then, the preprocessed image is segmented according to the image processing algorithm to obtain non-overlapping regions. The image processing algorithm can be one or more of the following methods: binarization, contrast thresholding, pixel gradient change, etc. Then, the segmented non-overlapping regions are subject to conditions such as area, perimeter, and shape. The regions that meet all the conditions are considered as valid feature regions in the real-time image. Finally, the feature region is compared with the standard image in the film thickness database to obtain the film thickness measurement data corresponding to the real-time image. In addition, the physical area of ​​the valid feature region is also calculated by the image processing unit to determine the actual area size of the valid feature region.

[0061] The above embodiment is adopted because, in practical use, due to the complex internal structure of the cavity, it is difficult to ensure that the cavity surface of the acquired image is on the same plane. This leads to differences in texture or color in different regions of the acquired real-time image. If only a fixed region in the image is compared with a standard image, the comparison results may be inaccurate. To solve this problem, this embodiment first preprocesses the received real-time image through an image processing unit to reduce interference from impurities, dust, or noise in the image on subsequent image comparisons. Then, the preprocessed image is segmented into multiple non-overlapping regions. These regions are then filtered using conditional constraints to identify truly effective feature regions. Finally, the multiple effective regions are compared with standard images in the film thickness database to obtain measured film thickness data for different regions under the real-time image, improving the accuracy of the comparison results. Furthermore, this embodiment also obtains the area size of the effective feature regions, further improving the accuracy of the data.

[0062] The above descriptions are merely embodiments of the present invention, and common knowledge regarding specific structures and characteristics is not elaborated upon here. It should be noted that those skilled in the art can make various modifications and improvements without departing from the structure of the present invention, and these should also be considered within the scope of protection of the present invention. These modifications and improvements will not affect the effectiveness of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. An online film thickness detection method for a cavity wax injection system, requiring the use of a detection system, the detection system comprising an image acquisition unit and an image processing unit, the image processing unit being connected to the image acquisition unit, characterized in that: The detection system also includes a driver, and a guide joint is connected between the output end of the driver and the image acquisition unit. The guide joint is used to drive the image acquisition unit to form different angles relative to the output end of the driver. The image acquisition unit includes an optical lens, an illumination source, and an ultraviolet light source. The image acquisition unit can acquire images of different areas inside the cavity. The image processing unit has a film thickness database, which includes several standard images corresponding to changes in film thickness while keeping the proportion of fluorescent agent added to the spray wax constant and the shooting angle constant. The film thickness database also includes several standard images corresponding to the proportion of fluorescent agent added to the spray wax and the change of shooting angle while the film thickness remains unchanged. The image processing unit can import the geometric data of the cavity and can determine the shooting angle when the image acquisition unit acquires the image based on the geometric data of the cavity and the position of the image acquisition unit in the cavity. It also includes a control unit, which is connected to the image processing unit, the driver, and the image acquisition unit; It also includes the following testing steps: Step A: The control unit checks whether the interlock signals between the driver, image acquisition unit, guide joint, and image processing unit are OK, and obtains the relevant position information; Step B: The control unit controls the driver to move the image acquisition unit into the wax injection cavity until the image acquisition unit moves to the set position; then the illumination source and ultraviolet light source of the image acquisition unit are turned on to provide conditions for the optical lens to acquire images; Step C: The control unit controls the guide joint to swing to adjust the shooting angle of the image acquisition unit. The control unit controls the image acquisition unit to perform image acquisition after each shooting angle adjustment. Step D: The image acquisition unit transmits the acquired image to the image processing unit in real time. The image processing unit compares the received real-time image with the standard image in the film thickness database to obtain the film thickness measurement data corresponding to the real-time image. The image processing unit stores the film thickness measurement data and sends the film thickness measurement data to the control unit. Step E: The control unit compares the received film thickness measurement data with the preset allowable data, and outputs an NG signal and issues an alarm signal when the film thickness measurement data exceeds the preset allowable data.

2. The online film thickness detection method for the cavity wax injection system according to claim 1, characterized in that: Before comparing the real-time image with the standard image in the film thickness database, the image processing unit needs to perform image preprocessing on the received real-time image.

3. The online film thickness detection method for the cavity wax injection system according to claim 2, characterized in that: Image preprocessing includes denoising and enhancement of the real-time image. The image processing unit uses image processing algorithms to segment regions with different textures and colors into non-overlapping regions. The image processing unit imposes conditional constraints on the segmented regions, and regions that meet all the conditional constraints are considered valid feature regions. Finally, the image processing unit compares the obtained valid region features with standard images in the film thickness database to obtain the measured film thickness data corresponding to the real-time image.

4. The method of claim 3, wherein: The image processing unit calculates the pixel area of ​​the effective feature region and, in conjunction with the nonlinear calibration result of the optical lens, converts the pixel area of ​​the effective feature region into a physical area in millimeters, thereby determining and outputting the size of the effective feature region.

5. The method of claim 3, wherein: The constraints include at least one of area constraints, perimeter constraints, and shape constraints.

6. The method of claim 3, wherein: The image processing algorithm is one or more of binarization, contrast thresholding, and pixel gradient transformation methods.

7. The online film thickness detection method for the cavity wax injection system according to claim 1, characterized in that: The guide joint is a physical control rod. The angle swing of the guide joint is controlled by connecting the physical control rod to the control unit.

8. The online film thickness detection method for the cavity wax injection system according to any one of claims 1-7, characterized in that: The location information includes the location information of the wax injection hole and the location information of the image acquisition unit.

9. The method of claim 1-7, wherein: When setting the shooting angle, make the shooting lens perpendicular to the cavity surface of the image being captured.

10. The method of on-line film thickness measurement of a cavity waxing system according to any one of claims 1-7, wherein: The control unit is a PLC controller.

Images